AI Grid performance?

So, the possibility of finally testing a full grid has arrived. For me this is the real performance bechmark, and what will really tell you if you are in for a sweet 60fps. ride in final, or a stuttering sub-par mess.

So, with a grid size of 20+ vehicles, starting at the back (all visible), what FPS are you getting and what is your hardware? Settings? GPU/CPU usage?

This will really tell how far the engine is in performance...

Personally i run it on a single GTX680 (since my SLI setup refuses to work) factory OC with a 4.2GHz. I5 CPU.

Runs smooth 60 with settings at the higher ranges.

And no, it is not playable below 60fps. No simulation is. :)
 
Also you will see how important good FPS is when racing online.
We had a good discussion about this in the staff section.
When a few guys were running Rfactor 2 at 30FPS and experiencing difficulties.
By getting the FPS higher this made the sim work better. @Marcel Offermans did explain it all very well in how one relates to another and what problems can be encountered in online racing
he also recommended no lower than 60FPS.
I myself had thought from what id been told 30 was good enough.
 
i just changed my screen to 50hz and run ac also in 50 hz and i could not tell any difference ... i read somewhere that AC physics also run at 50hz so i will just stay with it ... just gained 10 fps :)

EDIT: with faster cars like Lotus 49 i notest some difference on some turns ... not so smooth ....
 
Last edited:
I myself had thought from what id been told 30 was good enough.
It is and it isn't, depending. The human eye just works differently and every person is different. 30 FPS might work fine for the majority of people but there's always going to be people that can see through the effect. I have a friend who's colour blind and he says camouflage doesn't work on him. Where as my eyes would be tricked by the camouflage effect, the guy in camouflage would stand out for my colour blind friend.

There's also a lot of processing going on inside the brain, the image that comes off our eye is actually really poor, mostly blurry, has a big black spot in the middle a tiny spot beside that that's actually in focus and of course the whole image is upside down. Most of the picture you see is just the brains representation of data coming from the eyes more than a direct feed of what the eye sees, so there's a lot of scope for the brain to make stuff up as it's making the image. So I would guess that if a person thinks 60FPS is better than 30FPS it can have an effect on their interpretation of what they're seeing and their brain can actively look out for abnormalities in the 30FPS image which just compounds the problem.
 
... i read somewhere that AC physics also run at 50hz so i will just stay with it

That is wrong. These sims (AC, rF2, iRacing, ...) typically run their physics at something close to 400 Hz, so way higher than any frame rate and independent of it.

Think about it for a second, if the physics were linked to your frame rate, your car would behave differently depending on the frame rate you'd achieve. That would never give everybody a level playing field.

The reason they run at around 400 Hz is that some of the algorithms that are being used just would not be stable at 30-60 Hz.

But there are a few reasons why higher refresh rates are good to have, even if you claim you don't see much difference. The most important one is "lag". The slower the refresh rate, the bigger the delay before you actually see what happens after it has happened. And that means you will be slower to react, for example to your car loosing grip, and your chances of correcting such situations go down. Now that lag I'm talking about is the sum of many factors (input lag on your controllers, the time the computer takes to calculate positions, building the display, sending it to the monitor, etc..) so there are a lot of things you need to get right to end up with a better simulation in that regard.
 
I dont know much about this ... it was just what i read here http://www.assettocorsa.net/forum/i...jittering-motion-when-rendering-at-60hz.3756/

It may be wrong ... but a lot of people seem to agree with ... but like i said ... i dont know much about it and will not discuss it

Let me just state that it is highly unlikely. So far all the physics calculations I know about require much higher frequencies to be stable. Some people do point that out in the thread you refer to but as usual in forums, there's a lot of uninformed opinions being taken for the truth.

Some information about this:
Now one issue you might run into is how to hand over data from the much faster physics loop to the graphics loop. Judging from that original thread, it seems AC might be running at 250 Hz. Let's assume that is true, then if your graphics are refreshing at 60 Hz (let's say you've locked the framerate) then you might have an issue depending on how and when data is being handed over from physics to graphics. Again, assuming AC would do that only every 5th calculation, so at 250 / 5 = 50 Hz the two might be so out of sync that you can notice this effect. Then again, I have a high regard for the team behind AC, so I think they're much smarter than what I just came up with. So ultimately, let's see what they say, if they have time to answer all these questions. :)
 
What you say makes sense Marcel. I would definitely keep calculations on a (much) higher frequency than the actual display rate. Has been my aim for all computer apps I ever created (and those are nowhere near a racing sim). So, Jorge, it is quite interesting that you have a 10+ fps increase for the display, but it is getting too complicated for a layman as myself (too) to understand if the refreshrate affects your "feel" of the physics calculations. Are you sure it is not placebo? Anyway, will be fun to hear if KS have any say about it eventually.
 
So, Jorge, it is quite interesting that you have a 10+ fps increase for the display, but it is getting too complicated for a layman as myself (too) to understand if the refreshrate affects your "feel" of the physics calculations. Are you sure it is not placebo?

when i said 10+ FPS its because i run at 50fps instead of 60 ... i dont get more fps but i can set video settings a bit high without notest any fps drops ... thats what i mean with 10+ fps

changing from 60fps to 50fps didnt change anything on my driving feel (or physics perception)... i think you missundertood me :) running 50fps its exactly the same as 60fps for me except i can raise video settings that make the cars look a bit better (not much). Here and there i get the feeling that image is not so smooth but maybe thats the placebo effect cause im just expecting that something goes wrong on 50fps ... but i can really see no difference

but you just try for your self ... just remind that if you change to 50hz in AC you must also change it in your videocard settings or you will get lag
 
If 30 fps is enough and human eye cant see the difference above that,then how come all YouTube videos dont look smooth at 30 fps ?
Because we are indeed able to " see" above 30 easily...

The human eye and its brain interface, the human visual system, can process 10 to 12 separate images per second, perceiving them individually.[1] The threshold of human visual perception varies depending on what is being measured. When looking at a lighted display, people begin to notice a brief interruption of darkness if it is about 16 milliseconds or longer.[2] When given very short single-millisecond visual stimulus people report a duration of between 100 ms and 400 ms due to persistence of vision in the visual cortex. This may cause images perceived in this duration to appear as one stimulus, such as a 10 ms green flash of light immediately followed by a 10 ms red flash of light perceived as a single yellow flash of light.[3] Persistence of vision may also create an illusion of continuity, allowing a sequence of still images to give the impression of motion. Early silent filmshad a frame rate from 14 to 24 FPS which was enough for the sense of motion, but it was perceived as jerky motion. By using projectors with dual- and triple-blade shutters, the rate was multiplied two or three times as seen by the audience. Thomas Edison said that 46 frames per second was the minimum: "anything less will strain the eye."[4][5] In the mid- to late 1920s, the frame rate for silent films increased to between 20 and 26 FPS.[4]

When sound film was introduced in 1926, variations in film speed were no longer tolerated as the human ear is more sensitive to changes in audio frequency. From 1927 to 1930, the rate of 24 FPS became standard for 35 mm sound film;[1] the film travels through the projector at a rate of 456 millimetres (18.0 in) per second. This allowed for simple two-blade shutters to give a projected series of images at 48 per second. Many modern 35 mm film projectors use three-blade shutters to give 72 images per second—each frame is flashed on screen three times.[4]
----------------------------------------------
Frame rates in video games refer to the speed at which the image is refreshed (typically in frames per second, or FPS). Many underlying processes, such as collision detection and network processing, run at different or inconsistent frequencies or in different physical components of a computer. FPS affect the experience in two ways: low FPS does not give the illusion of motion effectively and affects the user's capacity to interact with the game, while FPS that vary substantially from one second to the next depending on computational load produce uneven, "choppy" movement or animation. Many games lock their frame rate at lower but more sustainable levels to give consistently smooth motion.

The first 3D first-person game for a personal computer, 3D Monster Maze, had a frame rate of approximately 6 FPS, and was still a success. In modern action-oriented games where players must visually track animated objects and react quickly, frame rates of between 30 and 60 FPS are considered acceptable by most, though this can vary significantly from game to game. Modern action games, including popular console games such as Halo 3, are locked at 30 FPS maximum, while others, such as Unreal Tournament 3, can run well in excess of 100 FPS on sufficient hardware. Additionally some games such as Quake 3 Arena perform physics, AI, networking, and other calculations in sync with the rendered frame rate - this can result in inconsistencies with movement and network prediction code if players are unable to maintain the designed maximum frame rate of 125 FPS. The frame rate within games varies considerably depending upon what is currently happening at a given moment, or with the hardware configuration (especially in PC games). When the computation of a frame consumes more time than is allowed between frames, the frame rate decreases.

A culture of competition has arisen among game enthusiasts with regard to frame rates, with players striving to obtain the highest FPS possible, due to their utility in demonstrating a system's power and efficiency. Indeed, many benchmarks (such as 3DMark) released by the marketing departments of hardware manufacturers and published in hardware reviews focus on the FPS measurement. LCD monitors of today are built with three major refresh rate in mind. The most common is 60Hz, which can be used at any resolution without requiring high quality computer systems to render, and then 120Hz and 144Hz. The 120Hz standard also supports what is known as 'lightboost' technology in some monitors, where strobing lights behind the monitor reduce ghosting at high FPS rates.

Beyond measurement and bragging rights, such exercises do have practical bearing in some cases. A certain amount of discarded “headroom” frames are beneficial for the elimination of uneven (“choppy” or “jumpy”) output, and to prevent FPS from plummeting during the intense sequences when players need smooth feedback most.

Aside from frame rate, a separate but related factor unique to interactive applications such as gaming is latency. Excessive preprocessing can result in a noticeable delay between player commands and computer feedback, even when a full frame rate is maintained, often referred to as input lag.

Without realistic motion blurring, video games and computer animations do not look as fluid as film, even with a higher frame rate. When a fast moving object is present on two consecutive frames, a gap between the images on the two frames contributes to a noticeable separation of the object and its afterimage in the eye. Motion blurring mitigates this effect, since it tends to reduce the image gap when the two frames are strung together. The effect of motion blurring is essentially superimposing multiple images of the fast-moving object on a single frame. Motion blurring makes the motion more fluid for some people, even as the image of the object becomes blurry on each individual frame. Motion blur can also induce headaches when people play a game that requires concentration.[16]

A high frame rate still does not guarantee fluid movements, especially on hardware with more than one GPU. This effect is known as micro stuttering.

Source: Wikipedia

So this subject cant have a "perfect" fps definition or a fixed fps for everything from movies to games.
Its possible for some people to see the difference from 50 to 60 or even 70 and slightly above,and for others to see no change above 30 or 40 and so on.
This depends on what monitors we have,PC hardware and games or videos themself.
Any change in this area will give different end results for different people.

For me,yes i can "see" the difference from 40 to 50 and even 50 to 60 (slightly) but above that,from 60 above i notice no change in my view.70,80,90,100..nothing.
The only thing that changes as the fps go up is the "Visual imput lag" of the ingame steering wheel (Assetto Corsa) where above 80 fps even that "visual imput lag" for the steering wheel is gone.
After this step all the extra fps you can get is for confort of having no issue when we run full AI races or online multiplayer because of fps "eaters". :D
 
This are my fps with custom,low and high settings in AC.

Firstly my rig,the important stuf:
CPU AMD Athlon II x4 640 @ 3.00 GHz
RAM 8 GB
GPU GTX 760 @ 2 GB.
2 Monitors 1280x1024 @ 69 Hz /Dual view

Testing with Ferrari 458 @ Silverstone: Single,12 AI and (unofficial) 24 AI

Custom ingame settings:

ScreenShot072_zps82ed5442.jpg

ScreenShot074_zpsd46a7096.jpg

ScreenShot073_zps84ecbbbe.jpg

ScreenShot075_zps82fc8cf2.jpg


Results:
...........Single car...........................12 AI...................................24 AI
MysettingsSinglecar_zpsca2e5890.jpg
Mysettings12AI_zps80785018.jpg
Mysettings24AI_zps05040bc4.jpg


Minimum/off settings ingame
Results:

...........Single car...........................12 AI...................................24 AI
VideoMinSinglecar_zpse1f5719d.jpg
VideoMin12AI_zps0faff2ea.jpg
VideoMin24AI_zpscd242a77.jpg


Maximum/on settings ingame
Results:

...........Single car...........................12 AI...................................24 AI
VideoMaxSinglecar_zpsc8582517.jpg
VideoMax12AI_zps2f913acb.jpg
VideoMax24AI_zps232875f2.jpg


The above screens are taken at start grid position
 
Hi

my graphic card works 97-99% with 24 cars on grid.

But if there are 40 cars it works only with 70 to 75% capacity.

Where is the catch?

PC: Asus ROG Maximus X WiFi Intel i7 8700k Corsair HX750i G.Skill RipJaws V 16GB, DDR4-3200 Asus Rog Strix 1080Ti OC Samsung m2 Evo 970 500gb Corsair H115i PRO
 

Latest News

Are you buying car setups?

  • Yes

  • No


Results are only viewable after voting.
Back
Top